Published September 1, 2016

RESEARCH

Wheat Landraces Currently Grown in : Distribution, Diversity, and Use

A. Morgounov,* M. Keser, M. Kan, M. Küçükçongar, F. Özdemir, N. Gummadov, H. Muminjanov, E. Zuev, and C.O. Qualset

A. Morgounov, N. Gummadov, CIMMYT, P.K. 39 Emek 06511 ABSTRACT , Turkey; M. Keser, ICARDA, P.K. 39 Emek 06511 Ankara, From 2009 to 2014 a nationwide effort was made Turkey; M. Kan, M Küçükçongar, F. Özdemir, Bahri Dagdas Inter- to document, collect, conserve, and character- national Agricultural Research Institute, Ereğli Yolu Üzeri PK:125 ize wheat landraces grown by Turkish farm- Karatay, , Turkey; H. Muminjanov, FAO–SEC, Ivedik Cad. No. ers. Spike samples were collected from more 55, 06170 Ankara, Turkey; E. Zuev, Vavilov Institute, B. Morskaya 44, than 1600 farmers from 59 provinces, planted S. Petersburg, Russia; C.O. Qualset, Dep. of Plant Sciences, Univ. of as single-spike progenies, and classified into California, 95616 Davis, USA. Received 24 Mar. 2016. Accepted 26 species, subspecies, and botanical varieties June 2016. *Corresponding author ([email protected]). Assigned (or morphotypes). Altogether, 95 morphotypes to Associate Editor Toi Tsilo. were identified representing three species and Abbreviations: FAO, United Nations Food and Agricultural Orga- six subspecies: einkorn wheat (Triticum mono- nization; H¢, Shannon diversity index; IWWIP, International Winter coccum L.), emmer wheat [T. turgidum subsp. Wheat Improvement Program. dicoccon (Schrank) Thell.], cone wheat (T. tur- gidum subsp. turgidum), durum wheat [T. turgi- dum subsp. durum (Desf.) Husn.], bread wheat heat is an important crop in Turkey with the planted (T. aestivum L. subsp. aestivum), and club wheat area of >7 million ha and annual production exceeding [T. aestivum subsp. compactum (Host) Mackey]. W 20 Tg (http://faostat.fao.org/). Annual consumption of bread and Compared with a nationwide survey in 1920, these findings represent a loss of 50 to 70% other wheat products in Turkey exceeds 200 kg per capita and is of the diversity found in 1920, though in four one of the highest in the world. The presumed center of wheat provinces, little if any loss occurred. Based on origin and diversity is situated in the Fertile Crescent (Feldman, the Shannon diversity index (H¢) and number 2001), which includes part of present-day Turkey. Thus, the of morphotypes, the highest diversity for bread diversity of wheat and its wild relatives in Turkey has a global wheat was observed in , Konya, Iğdır, role in providing important genetic resources for wheat improve- Diyarbakır, and Tokat provinces and for durum ment. A comprehensive review of the history, characteristics, and wheat in Adana, Diyarbakır, and Hatay prov- use of wheat landraces in Turkey has been recently published by inces. Socioeconomic data indicated that land- Karagöz (2014). There have been several major collection expedi- race farmers are found mostly in remote moun- tions for wheat landraces beginning shortly after establishment tainous subsistence communities with very little of the Turkish Republic in 1923 with an expedition by the Rus- grain trade, small areas planted to wheat, and sian Vavilov Institute following a route of some 12,000 km in relatively simple production technologies. The key reasons famers continue to grow landraces in 1925 and 1926, which documented agricultural prac- are their grain qualities and adaptation to abi- tices and crops and collected >5700 crop samples including 291 of otic stresses. In situ conservation should be tar- wheat (Zhukovsky, 1927). The Russian expedition was assisted by geted at provinces with the highest morphotype diversity, with the rarest landraces, and with the Published in Crop Sci. 56:3112–3124 (2016). highest share of farmers growing landraces. doi: 10.2135/cropsci2016.03.0192

© Crop Science Society of America | 5585 Guilford Rd., Madison, WI 53711 USA This is an open access article distributed under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

3112 www.crops.org crop science, vol. 56, november–december 2016 Turkish scientist Mirza Gökgöl who conducted a nation- and Tajikistan (Husenov et al., 2015). The fact that very wide inventory of wheat crop in the early 1930s. Upon few farming communities in few countries still maintain his request, seed samples with short descriptions (land- wheat landraces provides an important justification for race name, species, planting season, irrigated or rainfed, their inventory, collection, and conservation. elevation, and share of wheat area occupied) were sent The International Winter Wheat Improvement Pro- to him from all administrative provinces and districts gram (IWWIP; www.iwwip.org), based in Turkey, devel- of Turkey. Overall, 2120 samples were collected. They ops germplasm for central and western Asia. The IWWIP were all grown at the Yesilkoy Experimental Sta- is a cooperative program between the Ministry of Food, tion (present-day Istanbul Ataturk Airport) and classified Agriculture, and Livestock of Turkey, CIMMYT, and according to botanical traits. His findings were summa- ICARDA, and has operated since 1986. Its breeding rized in two volumes titled Turkish Wheat (Gökgöl, 1935, activities are implemented through a multilocational net- 1939). These books are an invaluable source of information work in Turkey in close cooperation with the key wheat on the status of wheat production prior to industrialized breeding programs in the region. The main emphasis of agriculture and provide baseline data for wheat landraces IWWIP breeding is broad adaptation, disease resistance, coverage prior to introduction of modern cultivars. and grain quality. Annually, IWWIP germplasm is sent Since the 1930s, major collections were made in 1948 to cooperators throughout the central and western Asia by Harlan (1950), who collected 2121 landraces from 63 region and globally for evaluation and selection. More provinces, and in 1984 by three international teams across than 65 cultivars originating from IWWIP germplasm provinces where wheat landraces were threatened. Sev- have been released in the central and western Asia region. eral regional and local collections and studies of wheat The IWWIP initially turned its attention to wheat land- landraces were done recently (Damania et al., 1996; races as a source of drought tolerance to enhance adap- Karagöz, 1996; Qualset et al., 1997; Tan, 2002; Karagöz tation for moisture-limiting environments and started and Zencirci, 2005; Akçura & Topal, 2006; Giuliani et testing them in yield trials. Their superior performance al., 2009). There are more than 22,000 Turkish wheat under drought suggested expansion of this work and, landraces in ex situ collections worldwide (https://www. thus, IWWIP examined options for collecting seeds of genesys-pgr.org). Peak collection years were in 1948 (1917 landraces from farmers’ fields. After this work started, it accessions, Harlan collection), 1970 to 1972 (1485 acces- was realized that the scale of wheat landrace cultivation in sions); 1979 (1846 accessions), and 1984 (2515 accessions). the country far exceeded expectations, IWWIP opted to In general, the past diversity of wheat landraces is well implement a countrywide inventory. represented in ex situ collections including the Turkey After collections have been made, it is important to Gene Bank in Menemen (İzmir). Socioeconomic aspects describe the diversity they represent. There is a method- of wheat landraces in several were ology for describing wheat genetic resources using mor- studied by Brush and Meng (1998) and the results served phological and agronomic traits (International Board for as the basis of an extended discussion on Anatolian wheat Plant Genetic Resources, 1985). Genomic technologies landraces by Brush (2004). However, no systematic coun- for diversity analysis have gained popularity for wheat and trywide study of existing landraces has been done since were recently reviewed by Khan et al. (2014). For example, M. Gökgöl’s work in the 1930s. The concept of national Alsaleh et al. (2016) recently completed a detailed study of inventories of the landraces gained high popularity in a collection of Kunduru durum wheat landrace accessions Europe and was supported by EU-wide projects (Negri et from Turkey and showed how genomic information can be al., 2012). These inventories not only targeted identifica- used to understand landrace diversity. In the present study, tion of what is grown where, but also suggested differ- we used the taxonomic and botanical description approach ent options for in situ conservation and use of landraces. that was used by previous collectors: Zhukovsky and Gökgöl However, in the case of wheat there are very few landraces in the 1920s and 1930s. This approach is based on assign- remaining in farmers’ fields in Europe. In North America, ment of each sample to a species and a botanical variety (or there is a community of amateur wheat landrace advo- morphotype as referred in this paper) based on the follow- cates, and seeds can be obtained from several providers ing highly heritable traits: presence or absence of awns and (http://www.ancientcerealgrains.org; http://www.grow- their color; color and pubescence of glumes, spike density, seed.org). Wheat landraces are still grown in Asia, on a and grain color (Zuev et al., 2013). This system was initially relatively substantial scale in northwestern Iran (Kooch- developed by Körnike in the 1880s and then expanded and eki et al., 2008) and Afghanistan (Buerkert et al., 2006), applied to Russian wheats in the 1900s by Flaksberger, later but no recent inventories have been conducted. In 2012 expanded by Vavilov in the 1920s and 1930s, while describ- to 2014, surveys and collections supported and guided by ing his worldwide collections, and eventually summarized FAO and CIMMYT identified, collected, and character- by Dorofeev in the 1970s (Zuev et al., 2013). Though it ized wheat landraces in Uzbekistan (Baboev et al., 2015) never gained wide-scale popularity, it is routinely used at

crop science, vol. 56, november–december 2016 www.crops.org 3113 Vavilov Research Institute of Plant Industry (Russia) and Table 1. Names and characteristics of the main morphot- ypes of Triticum aestivum ssp. aestivum and T. turgidum by gene banks in Germany, Poland, and Czech Repub- ssp. durum. lic. Carvalho et al. (2009) successfully used combination Morphotype designation Status of of genomic approaches and botanical variety description to by species† morphotype traits characterize old Portuguese bread wheat cultivars. Mor- Glume photype approach was also recently used for description of T. aestivum T. turgidum Presence pubes- Glume Grain durum wheat diversity (Carmona et al., 2010). ssp. aestivum ssp. durum of awns cence color color The Turkish national wheat landrace survey was con- albidum candicans No No White White ducted from 2009 to 2014 by IWWIP, the Turkish Min- lutescens schechurdinii No No White Red istry of Agriculture, and the United Nations Food and alborubrum subastrale No No Red White Agricultural Organization (FAO) with the aim of docu- milturum stebutii No No Red Red leucospermum muticovalenciae No Yes White White menting the presence of landraces across the country and velutinum muticinazillennse No Yes White Red collecting them for characterization, conservation, and delfi muticitalicum No Yes Red White use. A comprehensive description of this work was pro- pyrotrix No Yes Red Red vided by Kan et al. (2015) and an analysis of the socioeco- greacum leucurum Yes No White White nomic aspects by Kan et al. (2016). erythrospermum affine Yes No White Red erythroleucon hordeiforme Yes No Red White MATERIALS AND METHODS ferrugineum murciense Yes No Red Red Landrace collection was accompanied by a socioeconomic meridionale valenciae Yes Yes White White survey of the farmers growing the landraces. The methodol- hostianum durum Yes Yes White Red ogy of collections and survey was developed at a workshop turcicum italicum Yes Yes Red White in early 2009 and was largely based on a strategy adopted by barbarossa aegyptiacum Yes Yes Red Red Brush and Qualset in Turkey in the early 1990s (Brush, 2004). † Additional traits can be used in morphotype characterization: compactoides will The details of the methodology have been described by Kan be added to the morphotype designation for a semidense spike; plants of T. aesti- vum with dense spikes will be classified as subspecies T. aestivum ssp. compac- et al. (2015). Considering the large territory of Turkey and a tum with a different suite of morphotype designations depending on the four traits; wheat production area exceeding 7 million ha, the develop- spikes with black awns will be additionally distinguished by the prefix pseudo- ment of the collection plan represented a significant challenge. added to the morphotype designation. The identification of areas of landrace cultivation was based ~80% of the collection sites. In the remainder of cases, the coor- on the knowledge and experience accumulated at the agricul- dinates were recorded for the village nearest the field site where tural research institutes of the Ministry of Food, Agriculture collections were made. The socioeconomic survey instrument and Livestock. There are 12 institutes involved in wheat breed- used during face-to-face interviews with the farmers included 30 ing and research with regional mandates covering the whole points focusing on the following topics: (i) farmer family informa- country. They identified collection and survey areas in their tion, (ii) wheat production details, (iii) marketing, and (iv) farmers respective regions. The collection and survey teams usually perception of their landraces. About 80% of the collections were included a breeder, an agronomist, and a socioeconomist. The accompanied by completed surveys. For the remaining 20%, basic teams communicated with provincial agricultural authorities to data about the farmer, including name, family, and total area in identify the districts within the provinces where the landraces which wheat was grown, were recorded. The collections and sur- were likely to be grown. The teams further communicated with veys began in 2009 in three provinces (Erzurum, , and district administrations to identify the areas and villages likely Konya) and continued progressively until 2014 as the wide-scale to maintain landraces. This chain of communication was very cultivation of landraces in Turkey was revealed. In total, 1630 effective and demonstrated that the areas of the landrace culti- fields were visited, and collections of spikes were made from 1448 vation are commonly known by the agricultural administration of them. Seed bulks were collected from 162 farm stores. The at the district level. The teams visited the sites for preliminary collections were made in 59 out of the 81 provinces of Turkey, appraisals to develop the final collection plan. demonstrating a wide-scale cultivation of wheat landraces. The collections and surveys were made in June to August, The samples were sent to Bahri Dağdaş International depending on the region and crop development, to observe land- Agricultural Research Institute in Konya for analyses of the race fields near or at maturity and to collect spike samples. The landrace morphotype composition. The samples were classified team’s objective at the village level was to collect representative to wheat species following van Slageren (1994) (http://www.k- samples of the diversity of wheat landraces found there. Spike state.edu./wgrc/Taxonomy/comptri.html) and to morphotype collections in the fields were made randomly following diago- following classifications used by Dorofeev et al. (1979) and nal directions at regular intervals. The objective was to collect at recently updated for bread wheat by Zuev et al. (2013). The least 100 spikes from each field visited. The sample of spikes was morphotype description is based on qualitative morphologic tagged by the geocoordinates of each field, the farmer’s name, spike and grain traits with their combination defining specific and the local name for the landrace. Each sample of spikes is con- groups (Table 1). Zhukovsky (1927), while analyzing collec- sidered to represent a random population from the field where tions from Turkey, identified a group of bread wheats that were it was collected. Seed bulks were also collected from a farmer’s intermediate between bread wheat and club wheat. It was clas- storage when available. Precise geocoordinates were recorded for sified as a different type called grex compactoidum. Gökgöl (1939)

3114 www.crops.org crop science, vol. 56, november–december 2016 Fig. 1. Main regions of wheat landrace cultivation in Turkey based on surveys and collections in 2009 to 2014. 1, southern coastal; 2, Aegean; 3, western Black Sea; 4, central Black Sea; 5, central Anatolia; 6, northeastern Anatolia; 7, southeastern Anatolia; 8, central– eastern Anatolia; 9, eastern Mediterranean.  = approximate collection sites. also identified these forms as a separate group. This study fol- names, species, subspecies, and morphotype diversity. The lowed this classification as well. The number of spikes of each physical landscape, natural borders, mountains, and valleys species and morphotype was counted and recorded. Approxi- were also taken into account. The names of the farmers’ mately 70% of the landraces were mixtures of different species or morphotypes. A spike sample was assigned to wheat species wheats revealed that very similar landraces from the same and morphotype based on the major morphotype component village or district might have different names or landraces representing 65% or more of the overall composition. Some belonging to different species would have the same name. collections were discarded if they were modern cultivars or The detailed distribution of landraces at the province and were repetitions from the same landrace and the same farmer. district level is presented by Kan et al. (2015). The fol- In total, 1372 samples were available for field evaluations lowing nine major landrace cultivation regions are identi- including 1034 samples accompanied by socioeconomic data. fied in Fig. 1, showing approximate collection sites, and After description and classification, the spikes were planted as in Table 2 where limited descriptive information is given: head rows in the fall of the collection year in Konya, Eskişehir, or Erzurum (totaling >20,000 head rows). During the season, 1. Southern coastal region includes the all head rows were again classified by morphotypes to confirm in Mersin, Konya, Karaman, and Antalya provinces, or correct initial classification. Field selection of the best head the Bey Mountains of Mu la, Burdur, and Isparta rows took place while maintaining diversity. Selected head ğ rows were bulked, and their progenies were tested in yield provinces, and mountains surrounding the Menderes trials for two consecutive years. Agronomic data have been River through the provinces of Denizli, Aydın, and collected through replicated trials on >1000 landrace selec- İzmir. Almost 60% of all the landraces in this region tions but not presented here. Shannon diversity index was are durum wheat with morphotype hordeiforme being calculated for each province and region separately for bread the most common. Bread wheat, including the grex and durum wheat as described by Jain et al. (1975) based on the compactoidum type, accounts for 33.7%. frequency of different morphotypes for each geographic area. 2. includes the western transitional zone Genetic erosion of wheat landrace diversity from the 1920s was between the central Anatolia Plateau and the Aegean evaluated by comparing the number of landraces, number of and Seas. The overall landscape is moun- morphotypes, their frequency as reported by Gökgöl (1939), tainous with average elevation below 1000 m asl. and as observed in the present study. The seeds of collected The frequency of durum wheat landraces was 13.5%. landraces and passport data have been deposited at the Turkish Bread wheat landraces were grown in every province National Gene Bank in Ankara. with a high diversity of names and morphotypes. 3. Western strictly follows the RESULTS AND DISCUSSION Köroğlu mountain range. Durum and bread wheat Wheat Landrace Cultivation Regions landraces on average are almost equally repre- in Turkey sented. Einkorn wheat was found in four provinces: Definition of the main areas of wheat landrace cultiva- Bolu (landraces Yazlık Iza and Kislik Iza); Karabük tion was based on similarity with respect to local landrace

crop science, vol. 56, november–december 2016 www.crops.org 3115 Table 2. Geographic regions of wheat landraces cultivation and proportion of main landraces and morphotypes by wheat species. Triticum turgidum T. aestivum ssp. durum† ssp. aestivum† Others‡ Species or Main landrace (%) Main landrace (%) subspecies Region Province No. % (morphotype) % (morphotype) % (landrace) 1. Southern Adana, Antalya, Burdur, 193 59.6 Sarı B. (35.2) 26.4 Göderedi (27.8) 14.0 TAAC (Sarı B.); coastal Denizli, Isparta, İzmir, (hordeiforme); Koca (greacum, TAC Karaman, Konya, Mersin, B. (31.4) (hordeiforme, erythrospermum); Koca Muğla leucurum) B. (26.4) (erythroleucon, greacum) 2. Aegean Afyon, Balikesir, , 96 13.5 Ak B. (46.1) (leucurum); 42.7 Ak B. (26.8) (greacum, 43.8 TAAC (Çalıbasan); Çanakkale, Kütahya, Sarı B. (23.0) (leucurum) erythrospermum); Köse TM Manisa, Uşak B. (14.1) (albirubrum) 3. Western Black Bartın, Bilecik, Bolu, 115 35.6 Sarı Kelle (46.3) 38.3 Sarı B. (15.5) 26.1 TM (Yazlık, Sea Düzce, Eskişehir, Karabük, (murciense, (ferrugineum, Kislik Iza); Kastamonu, Sinop hordeiforme); Sofu B. erythrospermum); TTD (Gernik) (19.5) (murciense) Kırmızı Pazarcık (13.3) (erythrospermum, ferrugineum) 4. Central Black , Kırıkkale, Ordu, 130 57.6 Üveyik B. (22.7) 26.9 Çam B. (41.5) 15.5 TTD (Siyez, Sea Samsun, Tokat, (hordeiforme); Sofu B. (ferrugineum, Mahsul B.); (leucurum) erythrospermum); Ak TM (Mahsul B.) Dimenit (9.7) (greacum, erythrospermum) 5. Central Adana, , , 96 9.4 Şahman (80.9) 87.5 Kamçı (34.1) 3.1 Anatolia Kırşehir, Nevşehir, Niğde, (hordeiforme) (albirubrum); Zerun (30.9) (delfii) 6. Northeastern Ağrı, Artvin, Bayburt, 228 1.7 – 84.2 Kırik (23.1 (albirubrum, 14.1 TAAC; Anatolia Bitlis, Erzurum, , delfii); Karakılçık (7.7) TAC (Topbaş) Gümüşhane, Iğdır, Kars, (pseudo-meridionale); Van Topbaş–(6.7) (delfii, erytroleucum) 7. Southeastern Batman, Diyarbakır, 144 61.1 Kırmızı B. (19.3) 26.3 Kırmızı B. (18.4) 12.6 TAAC (Bağacak); Anatolia Mardin, Şanlıurfa, Siirt, (murciense); Sorgül (ferrugineum); Sergun TTC Şırnak (17.2) ( murciense, (18.4) (erythrospermum) hordeiforme); Karakılçık– (12.4) (reichenbachii, valenciae) 8. Central– Adıyaman, Bingöl, Elazığ, 166 31.9 Siverek (26.4) 52.4 Aşurelik B. (19.5) 13.8 TAAC (Kırmızı B., eastern Kahramanmaraş, Malatya (hordeiforme); Ohlemaz (erythroleucon); Çirpuz Ak B.); TTT Anatolia (15.1) (africanum) (18.4) (erytroleucum); Ağ Buğdayı (14.1) (erythroleucon) 9. Eastern Adana, Gaziantep, Hatay, 114 68.4 Karakılçık (24.3) 22.8 Kelbuğday (36.0) 7.8 TAAC; Mediterranean Kilis, Osmaniye (reichenbachii, (albidum, lutescens); TAC (Amik B.) leucomelan); Alibayır Bozbuğday (15.0) (9.0) (africanum); Havrani (erythrospermum, (9.0) (leucurum) greacum) † B., Buğday (wheat). ‡ TAC, T. aestivum ssp. compactum; TAAC, T. aestivum ssp. aestivum grex compactoidum; TM, T. monococcum; TTD, T. turgidum ssp. dicoccon; TTT, T. turgidum ssp. turgidum.

(Kaplıca Buğdayı); Kastamonu (Siyez, Gernik), and 6. Northeastern Anatolia region is almost entirely Sinop (Catal Siyez, Gernik). composed of high mountains including the Pontic, 4. Central Black Sea region represents the northern Mescit, Aras, and Ala ranges. There are almost no transitional zone with high rainfall and is domi- durum wheat landraces here. Wheat production is nated by durum wheat, especially in the Amasya, challenging because of severe cold in winter and Kırıkkale, and Samsun provinces. Hulled einkorn short, hot summers. Bread wheat landraces Kırik, wheat (landrace Mahsul Buğday) and emmer wheat Karakılçık, and Topbaş are cultivated on relatively (Siyez) were collected in . large scales as a result of their specific adaptation and 5. has scattered, relatively excellent grain quality. high mountains and also includes the 7. Southeastern Anatolia region follows the southeast- region famous for its landscape. The landraces of ern Taurus Mountains and the valley of the Tigris bread wheat were predominant (87.5%) with a great River. Frequency of durum wheat exceeds 60%, diversity of morphotypes: albirubrum, albidum, erythro- with a high diversity of morphotypes dominated by leucum, and delfii.

3116 www.crops.org crop science, vol. 56, november–december 2016 murciense. Bread wheat landraces are mainly ferru- then a higher number of botanical varieties translates into gineum and erythrospermum morphotypes. higher diversity. The Shannon diversity index was calcu- 8. Central– extends from lated for data from the 30 provinces where the number Adıyaman to Bingöl provinces along the Euphrates of collections exceeded 15 as well as overall for each of River. Though it is close and similar to the previ- the nine regions (Table 3). For durum wheat, the highest ous region, the composition of the landraces is quite diversity of morphotypes and Shannon index were in the different. Bread wheat landraces collectively exceed southeastern Anatolia and eastern Mediterranean regions 60% of all landraces collected. The landraces are also with Diyarbakır (Shannon index H¢ = 1.88) and Adana relatively uniform with the dominant erythroleucum (H¢ = 2.15) provinces having the most diverse durum morphotype. Durum wheat landraces are present in wheat landraces. For bread wheat, the highest diversity Adıyaman and Malatya provinces. was found for (Aegean region) account- 9. Eastern Mediterranean region includes the hills and ing for 38 morphotypes with the highest Shannon index mountains along the border with . The share among all provinces (2.65) and with the highest diversity of durum wheat landraces is the highest among all within individual landraces: 5.5 components per landrace. regions (68.4%) with the great diversity of landraces The province of Manisa is followed by Konya (H¢ = 2.21), and morphotypes. There is no dominating mor- Iğdır (2.20), Diyarbakır (2.17), Tokat (2.17), Adıyaman photype for durum wheat. Bread wheat landraces, (2.07), and Bitlis and Adana (both 2.06). Considering the although not numerous, were also represented by variation of all wheat species and subspecies the highest club wheat and grown in all provinces. diversity was in the Aegean region (four species; H¢ = 3.01) followed by the eastern Mediterranean (two species; In general, wheat landraces were found in all the eco- H¢ = 2.84) and northeastern Anatolia (two species; H¢ = logical regions of Turkey except the European Thrace 2.75) regions. Though the Shannon index for the western region, where highly industrialized crop production is Black Sea and central Black Sea regions is not high (2.08 practiced. In Regions 1, 4, 7, and 9, durum wheat land- and 2.27, respectively), they host very rare emmer wheat races dominate; in Regions 2, 5, 6, and 8, bread wheat and einkorn wheat landraces. landraces are more frequent; and in Region 3, they are The diversity of the landraces in the provinces may about equal (Table 2). A limited number of landraces derive from numerous diverse collections with very high were not assigned to bread or durum wheat as they were intralandrace homogeneity. Alternatively, the diversity complete mixture of two species with almost equal rep- may come from highly mixed landraces and populations. resentation. It is difficult to estimate the area of landrace Each situation requires a different strategy for collecting plantings in Turkey. Their widespread presence and great and in situ conservation. The former would be treated as diversity does not translate into coverage of large areas. fixed cultivars of a self-pollinated crop, while the main- Karagöz (2014) estimated that wheat and barley landraces tenance of the second group requires attention to the bal- are grown on an area exceeding 0.55 million ha. The cur- ance of different components. Some landrace cultivars, rent study agrees with this estimate. The other important like the durum wheat Kunduru, originated as selections question is how many landraces are cultivated in Turkey. from landrace populations and have been grown subse- This is again difficult to estimate without using modern quently in relative uniformity (Alsaleh et al., 2016). The genomic tools. Even in the same province or region, phe- structure of wheat landrace diversity is different depend- notypically similar landraces have different names. The ing on the region and province. In Manisa province, a third important question concerns the extent that the cur- center of bread wheat landrace diversity, only one col- rent collection included all the wheat landraces in Turkey lection out of 56 was homogeneous and the others rep- or whether some additional surveys are needed. By no resented mixtures or populations (Table 3). In provinces means did this study visit all areas growing wheat landra- such as Samsun, Kahramanmaraş, and Malatya, the share ces in Turkey. However, we believe that the majority of of homogeneous landraces exceeded 60%, demonstrating landraces have been collected. that they had been maintained for a long time in rela- tive purity, much like modern cultivars. Inter- and intra- Diversity of Wheat Landraces Among landrace heterogeneity is important for agronomic per- Provinces and Regions formance and the overall wheat genetic diversity and its From the conservation perspective, it is important to iden- drivers need to be understood. A preliminary Kompetitive tify the areas with the highest genetic diversity and the allele-specific polymerase chain reaction–single nucleo- areas with the greatest danger of losing rare genetic diver- tide polymorphism based genetic diversity study of >2100 sity (Negri et al., 2012). There is a diversity of landraces bread wheat lines selected from the landraces has been at a regional or provincial level. Assuming that morphot- completed (Morgounov et al., 2015). The results of the ypes of different wheat species represent units of diversity, genomic analysis will be coupled with the morphotype

crop science, vol. 56, november–december 2016 www.crops.org 3117 Table 3. Genetic diversity of wheat landraces in provinces and regions as estimated by the number of landraces, number of morphotypes (NoM), Shannon diversity index (H¢), percentage of uniform landraces (% ULR), and the number of morphotypes per landrace (M/LR). Triticum turgidum T. aestivum ssp. No. of ssp. durum aestivum All species landraces No. of Region Province in province species NoM H¢ NoM H¢ NoM H¢ % ULR M/LR 1. Southern Antalya 30 2 6 0.87 13 0.87 19 1.56 3.3 4.4 coastal Karaman 42 2 3 0.48 13 1.66 16 1.86 30.9 2.5 Konya 59 2 4 0.29 17 2.21 21 1.97 25.4 3.1 Mersin 25 2 4 0.67 7 1.58 11 1.41 28.0 2.3 Region 183 3 9 0.71 25 2.17 35 2.02 31.1 2.8 2. Aegean Manisa 56 3 7 0.58 38 2.65 47 2.75 1.8 5.5 Region 97 4 8 0.86 46 2.83 57 3.01 15.5 4.6 3. Western Bilecik 20 2 3 0.47 8 1.51 11 1.08 5.0 3.1 Black Sea Bolu 21 3 4 0.81 9 1.80 14 1.37 10.0 3.2 Sinop 23 3 3 0.40 5 0.52 9 1.43 26.1 2.4 Region 115 4 4 0.73 19 1.55 26 2.08 13.9 2.9 4. Central Samsun 30 4 3 0.29 8 1.74 14 0.83 70.0 1.6 Black Sea Tokat 56 2 3 0.75 30 2.17 33 2.27 19.6 3.1 Yozgat 16 2 4 0.92 9 1.27 13 1.73 25.0 3.2 Region 130 4 5 0.97 33 2.17 41 2.27 32.3 2.8 5. Central Aksaray 32 2 2 0.42 16 1.76 18 1.99 6.2 3.5 Anatolia Niğde 25 2 2 0.69 12 0.95 14 1.07 4.0 3.4 Sivas 19 1 0 – 7 0.48 7 0.48 10.5 2.3 Region 108 2 4 0.46 20 1.67 24 1.93 13.9 2.8 6. Northeastern Ağrı 56 2 1 – 19 1.77 20 1.78 1.9 3.2 Anatolia Bitlis 19 2 0 – 17 2.06 17 2.06 5.5 3.1 Erzurum 47 1 0 – 16 1.57 16 1.57 21.3 2.5 Giresun 30 2 3 1.01 13 1.65 16 1.87 16.7 3.0 Gümüşhane 19 1 0 – 12 1.86 12 1.86 47.3 2.1 Iğdır 33 2 2 0.06 22 2.20 24 2.29 36.4 2.2 Region 176 2 5 1.24 39 2.67 43 2.75 29.0 2.7 7. Southeastern Diyarbakır 59 2 11 1.88 20 2.17 31 2.69 33.9 2.2 Anatolia Mardin 39 2 3 0.44 5 0.78 7 1.15 38.5 2.0 Siirt 25 2 3 0.90 8 1.46 11 1.90 28.0 3.2 Region 144 2 11 1.71 22 1.95 33 2.48 31.2 2.4 8. Central– Adıyaman 27 2 9 1.24 16 2.07 28 2.35 22.2 2.9 eastern Elazığ 16 2 1 – 10 0.64 12 1.08 68.7 2.2 Anatolia K. maraş 25 2 1 – 10 1.55 11 1.70 62.5 2.0 Malatya 95 2 5 1.13 18 0.94 25 1.70 60.0 1.9 Region 165 2 9 1.20 32 1.35 46 2.00 49.1 2.1 9. Eastern Adana 28 2 11 2.15 16 2.06 27 2.77 25.0 3.1 Mediterranean Gaziantep 15 2 4 1.12 7 1.25 11 1.48 53.3 2.0 Hatay 45 2 10 1.71 16 1.41 26 2.25 35.3 2.4 Region 114 2 14 2.21 27 2.32 41 2.84 35.9 2.4 Overall 1229 4 14 1.69 64 2.68 89 3.06 29.5 2.9 classification and agronomic data to delineate individual diversity collected in this study. One of the challenges in landraces and establish a core set of germplasm for further making such a comparison was the change in the admin- detailed study. istrative division of provinces and districts as well as their renaming, which took place in the last 80 to 90 yr. While Erosion of Wheat Landraces Diversity in the Gökgöl books listed 60 provinces, present-day Turkey 2010s versus 1920s is divided into 81 provinces. The Gökgöl surveys iden- Gökgöl (1935, 1939) provided detailed botanical descrip- tified the actual area cultivated with different landraces, tion of the landraces following Vavilov’s approach. The while the current study used the frequency of the landra- Gökgöl data were transferred into spreadsheets and served ces collected with the assumption that it could be appro- as a baseline for evaluation of the changes in landraces priately translated into an area estimate.

3118 www.crops.org crop science, vol. 56, november–december 2016 Table 4. Wheat landrace diversity for selected provinces in Turkey found in the 1920s compared with the current results (2000s). Percentage of landraces belonging No. of to different species‡ morphotypes Province No. of Per Region (districts)† Period landraces TAA TAC TTD Others Total landrace Southern Adana (Feke, 1920s 8 12.8 7.4 50.4 29.4 (TTT) 31 3.9 coastal Kozan, Saimbeyli) 2000s 10 47.4 3.1 45.1 0 35 2.7 Konya (Bozkır, Hadim) 1920s 6 2.8 48.8 48.4 0 20 3.3 2000s 8 35.5 5.7 57.9 0 22 2.7 Aegean Manisa (Akhisar, Kula) 1920s 9 5.9 24.4 69.7 0 47 5.2 2000s 10 78.8 0.4 17.0 3.8 (TTT) 41 4.1 Western Black Bolu (Mudurnu) 1920s 5 71.1 10.0 18.7 0 16 3.3 Sea 2000s 4 11.5 0 29.1 54.4 (TM) 4 1.0 Central Black Samsun (Ladik, 1920s 11 22.0 2.2 75.8 0 29 2.6 Sea Vezirköprü) 2000s 4 1.5 0 81.3 17.2 ( T T D,T M) 16 4.0 Tokat (Merkez) 1920s 5 33.5 10.6 55.9 0 22 4.4 2000s 9 39.0 0.1 58.3 0 26 2.9 Yozgat (Merkez, 1920s 2 2.8 19.5 77.7 0 17 1.5 Akdağmadeni) 2000s 5 40.5 0.5 57.4 0 13 2.6 Northeastern Ağrı (Diyadin, Eleşkirt) 1920s 9 97.4 0.4 0.1 2.12 (TTC) 38 4.2 Anatolia 2000s 4 100.0 0 0 0 7 1.7 Erzurum (Oltu, Pasinler) 1920s 16 84.4 1.8 10.7 2.3 (TTC) 68 4.2 2000s 6 75.3 23.0 0 0 13 2.2 Giresun (Şebinkarahisar) 1920s 5 23.1 43.2 33.6 0 26 5.2 2000s 9 83.6 0 9.4 0 13 1.4 Gümüşhane 1920s 10 41.1 28.0 31.9 0 33 3.3 (Kelkit, Torul) 2000s 5 100.0 0 0 0 12 2.4 Southeastern Diyarbakır (Lice, 1920s 12 64.8 0.1 35.8 0 27 2.2 Anatolia Çermik, Ergani) 2000s 10 33.8 3.2 59.9 0 12 1.2 Mardin (Midiyat, Savur) 1920s 5 10.3 21.7 67.9 0 23 4.6 2000s 3 28.4 0 71.5 0 4 1.3 Siirt (Eruh, Merkez) 1920s 5 51.8 7.8 39.5 0 34 6.8 2000s 7 42.0 0.2 57.8 0 21 3.0 Central–eastern Elazığ (Baskil, 1920s 13 55.8 5.7 27.7 5.2 (TTT) 32 2.5 Anatolia Merkez, Palu) 2000s 3 87.1 1.1 0 11.2 (TTT) 7 2.3 Kahramanmaraş 1920s 5 42.9 0.9 38.5 17.6 (TTT) 22 4.4 (Elbistan) 2000s 2 24.5 0 72.6 0 2 1.0 Malatya (6 districts) 1920s 20 56.0 9.7 23.9 0 46 2.3 2000s 10 69.6 0.6 29.3 0.1 (TTT) 22 2.2 Overall 1920s 146 39.9 14.2 41.5 3.3 213 3.7 2000s 109 52.9 2.2 38.0 5.1 63 2.3 † Merkez designates the central district of each province. ‡ TAA, T. aestivum ssp. aestivum; TAC, T. aestivum ssp. compactum; TTC, T. turgidum ssp. carthlicum; TTD, T. turgidum ssp. dicoccon; TTT, T. turgidum ssp. turgidum; TM, T. monococcum.

For a combination of 17 provinces and districts, the Kırik (cracked) at 6.7%, and Kırmızı Buğday (red wheat) at coverage of Gökgöl’s surveys was matched by the cur- 4.1%. As agronomy evolved, climate changed, and adaptive rent study (Table 4). The first observation was that the requirements for wheat production changed, the landraces names of the landraces changed over time almost entirely. evolved and acquired new names reflecting their proper- Gökgöl mentioned generic names of 93 landraces, while ties, use, or origin. The second observation was that in the the current collection from the same provinces and dis- past, up to 40% of wheat was planted in the spring, and tricts mentioned 81 names. However, only 18 names were currently, >95% of the landraces are planted in the fall. preserved. In the 1920s the most common name was Kislik In the 1920s, throughout the provinces and districts (winter wheat) at 15.6%, followed by Karakılçık (black- listed in Table 4, bread wheat and durum wheat occu- awn wheat) and Yazlık (spring wheat) at 8.0% each. Cur- pied almost equal acreage: 39.9 and 41.5%, respectively, rently the most common name is Ak Buğday (white wheat) with an additional 14.2% of area devoted to club wheat. at 9.8% followed by Sarı Buğday (yellow wheat) at 7.9%, Currently, bread wheat landraces dominate planted

crop science, vol. 56, november–december 2016 www.crops.org 3119 areas (52.9%), durum wheat landraces have been slightly Table S1 and for selected provinces in Table 5. The land- reduced (38.0%), and club wheat landraces almost entirely race farmers are found mostly in remote mountainous disappeared (2.2%). This tendency of higher frequency subsistence communities with market distance exceeding of bread wheat at the expense of club and durum wheat 15 to 20 km and having very little grain trade (<3.4% is more pronounced in Adana, Konya, Manisa, Yozgat, for bread wheat and <6.4% for durum wheat). Cultivated Giresun, Gümüşhane, and Elazığ provinces. The fre- wheat area per farm varied from <1 ha in the southern quency of durum wheat landraces increased in areas tradi- coastal region to a maximum of 4 to 5 ha in the central tionally growing this crop: Diyarbakır, Mardin, and Siirt. Black Sea region, which is lower than regional averages. The changes in environments, agronomy, and farmers’ The share of farmers never changing seed exceeds 43% for preferences resulted in substantial changes affecting not bread wheat landraces and 31% for durum wheat landra- only the proportion of different species but their composi- ces. Hand planting or use of primitive spreaders is prac- tion as well. The predominant morphotypes had changed ticed, on average, by 74% of farmers and hand harvesting in 30% of the districts for bread wheat landraces and in by almost 40%. A substantial share of farmers apply seed 50% of the districts for durum wheat landraces. chemical treatment to prevent seed-transmitted diseases. The loss of genetic diversity can be estimated by the The average age of farmers exceeds 53 yr and 90% of frequency of rare species. Cone wheat disappeared in them either lack formal education or graduated only from Adana and Kahramanmaraş but was observed in Manisa primary school. On average, there are >5.3 people per province after not being registered there before. Einkorn household growing landraces vs. the national average for wheat was not observed in Bolu and Samsun provinces rural areas of 4.9. Depending on the region, up to 80% of in the past, but several fields of it were identified by this the farmers have tried modern cultivars and most of them study. The reduction in the number of morphotypes per kept growing them along with landraces. The proportion province varied from 90% in Kahramanmaraş to an actual of area growing wheat landraces to total wheat area in increase in . Erzurum had the highest farmers’ fields varied from 45 to 55% in the central Black number of morphotypes in the past (68) compared with Sea region and up to 98% in the southern coastal region. the 13 identified now. Manisa province lost very little— Farmers have access to modern cultivars but still keep only six morphotypes. Adana, Konya, and Tokat prov- their landraces. The main reason for maintaining landra- inces seemed to gain some diversity. The landraces grown ces is satisfaction with the landraces’ performance. While, now are more homogeneous (on average 2.3 morphotypes on average, only 25 and 30% (bread wheat and durum per landrace) than the observations of the 1920s (3.7). The wheat growers, respectively) of the farmers rated yield of total number of morphotypes listed by Gökgöl for the the landraces as good; 83% of the respondents for bread provinces and districts in Table 4 was 213 vs. 63 identified wheat and 93% for durum wheat were happy with the in the present study, a loss of 70%. However, the descrip- grain quality and its suitability for homemade products tion of the morphotypes by Gökgöl went into minor (Fig. 2). The other highest ranked traits for bread wheat details such as the length of awns (using the prefixsub- for and durum wheat, respectively, were straw yield (74 and morphotypes with shorter awns), the color of the glume 80%) and straw quality (70 and 76%), cold tolerance (78 (using prefixes: griceo-, negro-, triste-, rubro- for different and 82%), and drought tolerance (71 and 84%). For most colors), and the state of the culm (using prefix pleno- for of these traits, durum wheat landraces were rated slightly solid culm). This detailed description by Gökgöl identi- higher than bread wheat landraces. It appears that excel- fied 62 minor morphotypes with frequencies below 0.2%. lent adaptation of landraces to cold and drought coupled In the present study, such a detailed degree of description with highly suitable grain quality were the key factors was not attempted because of limitations in capacity and considered by farmers when they decide to keep the land- time. However, even if we remove 62 minor morphotypes races. The second, less obvious, but also important, factor from the estimate of Gökgöl, there has still been a 59% is tradition and inclination of the farming communities loss of distinct morphotypes. If we compare all collections to follow the practices of the older generation and peers. of the 1920s with those of the present time without a focus Very substantial differences between provinces in the on the common comparison areas of Table 4, from all the diversity of wheat landraces (Table 3, 5) may be explained eight species or subspecies, 388 morphotypes were identi- by diversity in the wheat production environments, fied then compared with 95 now, a 75.5% reduction. changeable weather and climate, different farmers’ prac- tices (including use of machinery and the end use of grain The Drivers of Wheat Landrace Diversity and straw), initial wheat diversity at the time when indus- The first and foremost question is why Turkish farmers still trial agriculture started, rate of spontaneous hybridization, cultivate wheat landraces. Socioeconomic data collected and possibly other factors. Characteristics of the farmers in during the survey and summarized by regions for durum the provinces with the high (Manisa, Tokat, Diyarbakır, and bread wheat growers is presented in Supplemental and Adana) and relatively low (Samsun, Mardin, and Siirt)

3120 www.crops.org crop science, vol. 56, november–december 2016 Table 5. Socioeconomic and technical characterization of farmers growing wheat landraces in selected provinces of Turkey.

Region Eastern Aegean Central Black Sea Southeastern Anatolia Mediterranean Selected province Parameter Manisa Samsun Tokat Diyarbakır Mardin Siirt Adana No. of farmers surveyed 48 21 53 57 30 25 28 No. of wheat species 3 4 2 2 2 2 2 No. of morphotypes 47 14 33 31 7 11 27 Shannon diversity index (H¢) 2.75 0.83 2.27 2.69 1.15 1.90 2.77 No. of morphotypes per landrace 5.5 1.7 3.2 2.1 1.9 2.7 3.1 population Elevation above sea level (m) 667 735 1074 831 968 826 1062 Farmer age, years 53.5 51.4 55.2 54.1 48.3 43.0 49.3 No. of people in household 4.7 5.0 4.1 8.2 6.4 4.6 3.8 Market distance (km) 21.8 17.4 17.7 19.0 14.6 36.1 28.1 Total land (ha) 5.7 9.6 5.8 5.9 5.5 4.2 2.6 Total wheat area (ha) 1.3 6.2 3.5 3.3 2.4 1.8 1.2 Share of landraces in total wheat (%) 100 28.2 52.9 73.7 100 100 84.0 Percentage of farmers: Without education or with primary 95.8 95.4 90.6 90.2 100 100 92.3 school only Using wheat for bread 64.6 0 18.9 18.9 0 0 46.4 Using wheat for bulgur 27.1 0 26.4 14.0 0 0 46.4 Using their own seed 97.9 85.7 84.9 89.4 100 94.4 92.8 Never changing seed 79.2 57.1 5.7 0 0 0 78.6 Cleaning seed 82.1 80.9 73.6 91.2 3.3 5.5 82.1 Trying modern cultivars 0 100 64.1 52.6 3.3 0 21.4 Keeping modern cultivars 0 90.5 58.5 50.9 3.3 0 21.4 Planting by hand or spreader 100 80.9 86.8 89.5 76.7 72.2 100 Harvesting by hand 75.0 42.9 32.1 17.5 3.3 0 75 Selling grain 0 9.5 1.2 1.7 0 0 0 Selling straw 16.7 85.7 24.5 5.2 0 0 0

Fig. 2. Percentage of farmers’ ratings of different traits of bread wheat (BW) and durum wheat (DW) landraces as good based on a survey of 1026 households in Turkey in 2009 to 2014. diversity of landraces are presented in Table 5. It may grow very diverse landraces. In Manisa province, none of appear that older farmers from Diyarbakır province grow the farmers grow modern cultivars; they only grow land- more diverse landraces than relatively younger farm- races, and the diversity is very high. On the other hand, ers from Mardin and Siirt provinces in the same region. in Tokat and Diyarbakır provinces, >50% of the farmers However, relatively young farmers of also grow diverse landraces and modern cultivars by side.

crop science, vol. 56, november–december 2016 www.crops.org 3121 It seems that there is no clear socioeconomic premise that Table 6. Socioeconomic characteristics of farmers growing only one wheat landrace versus farmers growing two or more explains higher diversity in one province vs. the other. landraces. On-farm diversity of wheat is of fundamental impor- tance for sustainable production and diversity conserva- Farmers growing: tion. This diversity is manifested by farmer cultivation of Two or One more more than one landrace, heterogeneity vs. homogeneity Characteristic landrace landraces of landraces, or cultivation of a combination of landraces No. of farmers 1016 99 and modern cultivars. Kan et al. (2015, 2016) analyzed the Percentage of farmers 91.3 9.7 current data set and concluded that except for age, none No. of morphotypes per landrace population 2.79 3.00 of socioeconomic factors (number of people and the labor Farmers age (yr) 55.3 52.7 force in the households, educational level, etc.) affected Percentage of farmers graduating primary 68.8 73.2 school and above the farmers’ preferences for growing wheat landraces only No. of people in household 5.6 5.0 or both modern cultivars and landraces produced together. Total farm land (ha) 5.7 7.6 The share of farmers growing only landraces in the age Wheat area (ha) 2.6 3.3 group >50 yr old was 64.0%, while among the group of Share of landraces in wheat area (%) 80.4 83.1 farmers younger than 50 yr old only 36.0% of growers Percentage of farmers: opted for landraces only. This is another warning sign that who tried modern cultivars 36.6 31.0 the newer generation of farmers is more willing to try the who kept modern cultivars 33.3 28.1 modern cultivars, and data showed that up to 80% of those who use hand planting 64.2 65.9 who tried were likely to keep them in the farming system. who use own seed 85.5 94.8 The survey identified 99 farmers (9.7%) who culti- who never change the seed 39.4 44.6 vate more than one landrace (Table 6). These are possibly who use hand harvesting 41.3 38.9 Percentage of farmers rating the following traits as good for champions of wheat diversity who are keen to work with their landraces more landraces. Most of these farmers (65) maintain dif- Grain yield 23.5 32.2 ferent wheat species probably to meet different end-use Disease resistance 48.6 56.7 needs. The number of morphotypes found per landrace Cold tolerance 75.9 76.0 population in this group of farmers is 10% higher than that Drought tolerance 70.1 75.1 found for farmers who grow only one landrace. Interest- Grain quality 85.5 88.7 ingly, these farmers are slightly younger (52.7 yr vs. 55.3) and better educated, own substantially more land (7.6 ha and Meng, 1998). While considering the household deter- vs. 5.7 ha), grow more wheat (3.3 ha vs. 2.6 ha), rely more minants of wheat diversity, higher wealth, better education, on their own seed production (94.8 vs. 85.5%), and change and younger age were the main factors reducing on-farm seed less frequently than the group of farmers who grow diversity through incorporation of modern cultivars or cul- only one landrace. The degree of machinery use for plant- tivation of more uniform landraces. At the same time, a ing and harvesting is similar between the two groups. The high probability of the household to cultivate landraces was farmers growing more than one landrace are better satis- not related to the degree of diversity of individual landraces. fied with the yield performance (32.2 vs. 23.5%), disease resistance (56.7 vs. 48.6%), and drought tolerance (75.1 Home Use of Wheat Landraces vs. 70.1%) of their landraces. They are less likely to grow Wheat grain in the rural areas is used for two main pur- modern cultivars, thus preserving the landraces. These poses: bread, including typical loaves and thin types, farmers have been registered in the study database and are and bulgur or cracked wheat, which is cooked in water. potential entry points for programs to reinforce on-farm, Respectively, bread and durum wheat are normally used in situ conservation of wheat landraces. for these two products. Based on the survey of the farmers Overall, the underlying causes of differences in diver- in the regions growing primarily bread wheat (Aegean, sity among geographical regions and administrative prov- central Anatolia, northeastern Anatolia, and central–east- inces are still not clearly understood and require more in- ern Anatolia), its grain is mainly used for bread (64.3 to depth analysis. Effects of elevation and biotic and abiotic 83% of farmers; Table 7). Of the four regions dominated stresses play an important role. Damania et al. (1996) dem- by durum wheat, grain in the southern coastal and eastern onstrated that Turkish durum wheat landraces were more Mediterranean regions is mainly used for bulgur (55.5 and diverse at lower elevations. A survey of 287 households in 87.1%, respectively). The durum grain in the central Black Eskişehir, Kütahya, and Uşak provinces in 1992 demon- Sea and southeastern Anatolia regions is used for both strated that environmental heterogeneity and the risks and bulgur and bread (61.1 and 83.3%, respectively). Gener- high cost of obtaining wheat products with the desired ally, the farmers were quite flexible in dual use of their quality was the main reasons to maintain landraces (Brush grain for bread, bulgur, and other homemade products.

3122 www.crops.org crop science, vol. 56, november–december 2016 Table 7. Use of bread wheat (BW) or durum wheat (DW) land- CONCLUSIONS races in different regions of Turkey (percentage of farm- ers surveyed). Despite the small total area of wheat landrace production in Turkey, the nationwide landrace distribution and diversity End-product Landrace quality rated is significant for the global heritage of humankind, and it Region Bread Bulgur Both as good has now been well documented, collected, and preserved. ————————————— % ————————————— The reduction in number of morphotypes compared with 1. Southern coastal the 1920s was documented in this study and exceeded 70% DW 9.5 55.5 35.2 96.8 in some regions. However, comparing wheat landraces of BW 60.0 5.0 31.7 88.3 the past with those of the present demonstrated that they 2. Aegean are evolving and very different from what they were 90 to DW – – – 100 yr ago. This study suggests that long-term cultivation BW 83.0 3.8 13.2 75.0 of wheat landraces and their exchange by farmers results 3. Western Black Sea in their continuous enhancement, adaptation, and cre- DW 5.1 74.4 5.1 95.2 BW 45.2 33.3 4.8 97.4 ation of new genetic diversity. In situ conservation efforts 4. Central Black Sea in Turkey should focus on three major areas: (i) provinces DW 4.2 27.8 61.1 72.2 with the highest diversity of landraces as identified by the BW 55.5 14.8 37.1 31.5 number of morphotypes and Shannon diversity index 5. Central Anatolia (Adana, Adıyaman, Aksaray, Bitlis, Diyarbakır, Hatay, DW 10.5 84.2 5.2 100.0 Manisa, and Tokat provinces); (ii) provinces hosting rare BW 75.4 5.3 19.3 89.6 species like einkorn and emmer wheat (Bolu, Karabük, 6. Northeastern Anatolia Kars, Kastamonu, Kütahya, Samsun, and Sinop); and (iii) DW – – – provinces with the highest share of farmers growing both BW 64.6 0.5 28.7 86.1 landraces and modern cultivars because of the likelihood 7. Southeastern Anatolia of the latter replacing the former (primarily provinces of DW 2.4 14.3 83.3 96.3 the western Black Sea and central Black Sea regions). BW 41.6 8.4 50.0 92.8 8. Central–eastern Anatolia Future wheat landrace conservation and utilization DW 0 11.8 88.2 100 efforts in Turkey should be directed at two targets. The BW 64.3 5.1 26.5 92.6 first is how to conserve existing landraces and possibly 9. Eastern Mediterranean expand their area and diversity. In addition to important in DW 9.7 87.1 3.2 95.4 situ conservation practices and policies, genetic improve- BW 67.7 19.3 3.0 93.6 ment of landraces and returning them to farmers may be a viable option. Modern breeding tools allow rapid selec- Most of the club or compact wheat is used for dual pur- tion and incorporation of desired traits while preserving poses. Hulled einkorn wheat is used for bulgur in the Bolu the overall integrity of the landraces. Their competitive- region and for animal feed elsewhere. Emmer wheat is ness with modern cultivars could be improved and lead to almost entirely used for animal feed. greater and more enduring use by farmers. In 2015, the The degree of satisfaction with the grain quality was not International Treaty on Plant Genetic Resources (http:// related to the end use of the landraces. Durum wheat farm- www.planttreaty.org) awarded CIMMYT a project to ers in the central Anatolia region were 100% satisfied with conserve wheat landraces on farmers’ fields in Turkey, the grain, mostly using it for bulgur. In the southeastern Afghanistan, and Iran. The second target should be fur- Anatolia and central–eastern Anatolia regions, the durum ther description, classification, evaluation, and use in farmers also gave very high ratings to the quality of their breeding and research programs of the tremendous diver- landraces, using them for dual purposes (bread and bulgur). sity of the collected wheat landraces. Up to now, >1000 Bread wheat farmers in the western Black Sea region were landrace selections have been subjected to a robust system highly satisfied with their grain (97.4%), using it for bread of multilocational testing in Turkey, and superior geno- and bulgur. The lowest rating of grain quality was in the types have been identified. Having survived for so long, central Black Sea region: 72.2 and 31.5% for bread wheat these landraces are highly valuable genetic resources for and durum wheat landraces, respectively. This may be meeting the challenges of modern agriculture. explained by interference of modern cultivars in the rating because this region has the highest share of farmers growing Acknowledgments both landraces and improved cultivars. This region is also IWWIP is financially supported by the Ministry of Food, Agri- characterized by high rainfall, including during the wheat culture and Livestock of Turkish Republic and CGIAR–CRP maturation period, which may affect quality. WHEAT. FAO–SEC partly supported collections and surveys in 2012 and 2013. Dr. Kai Sonder is highly appreciated for pre- paring the map with landrace locations. Survey and collections crop science, vol. 56, november–december 2016 www.crops.org 3123 participants are highly recognized and sincerely thanked: Dr. Harlan, J.R. 1950. Collection of crop plants in Turkey. Agron. J. Enver Kendal, Dr. Hakan Hekimhan, Dr. Mustafa Güllü, Assis- 42:258–259. doi:10.2134/agronj1950.00021962004200050012x tant Prof. Köksal Karataş, Dr. Emin Dönmez, Sevinç Karabak, Husenov, B., M. Otambekova, A. Morgounov, and H. Muminjanov. Rahmi Ta çı, O uzhan Ulucan, Soner Yüksel, Mustafa Çakmak, 2015. Wheat landraces in farmers’ fields in Tajikistan: National sur- ş ğ vey, collection, and conservation, 2013–2015. FAO, Ankara, Turkey. Dr. Ümran Küçüközdemir, Enes Yakışır, İbrahim Kaçmaz, Ercan International Board for Plant Genetic Resources. 1985. Descriptors of Doğan, Erdal Kablan, Adem Eskici, Assistant Prof. Fatih Sarı, Prof. wheat (revised). IBPGR Secretariat, Rome, Italy. Dr. Ela Atış, Mert Altıntaş, Muhittin Böget, Süleyman Bakırcı, Jain, S.K., C.O. Qualset, G.M. Bhatt, and K.W. Wu. 1975. Geographic Murat Kan, Dr. Arzu Kan, and Assistant Prof. Sezen Küçükçon- patterns of phenotypic diversity in a world collection of durum gar. Dr. Patrick McGuire’s suggestions and editorial work with wheats. Crop Sci. 15:700–704. doi:10.2135/cropsci1975.0011183X0 the paper are highly appreciated. Open access to this paper was 01500050026x supported by International Treaty on Plant Genetic Resources Kan, M., M. Küçükçongar, M. Keser, A. Morgunov, H. Muminjanov, F. Özdemir, and C. Qualset. 2015. Wheat landraces in farmers’ for Food and Agriculture project W2B-PR-41-TURKEY with fields in Turkey: National survey, collection, and conservation, the financial assistance of European Union. This study would not 2009–2015. FAO, Ankara, Turkey. have been possible without the participation of the Turkish farm- Kan, M., M. Küçükçongar, A. Morgunov, M. Keser, F. Özdemir, H. ers growing wheat landraces, providing access to their fields, and Muminjanov, and C. Qualset. 2016. Wheat landraces production responding to the surveys and the village seniors and district and on farm level in Turkey: Who is growing where? Pak. J. Agric. Sci. province administrators who facilitated collections and surveys. 53:159–169. Karagöz, A. 1996. Agronomic practices and socioeconomic aspects of emmer and einkorn cultivation in Turkey. In: S. Padulosi, K. Ham- References mer, and J. Heller, editors, Proc. First Int. Workshop on Hulled Akçura, M., and A. Topal. 2006. Phenotypic diversity of winter bread Wheats. Castelvecchio Pascoli, Tuscany, Italy. 21–22 July 1995. Inter- wheat landraces in Turkey. J. Crop Res. (Turkey) 3:8–16. national Plant Genetic Resources Institute, Rome, Italy. p. 172–177. Alsaleh, A., F.S. Baloch, M. Nachit, and H. Özkan. 2016. Pheno- Karagöz, A. 2014. Wheat landraces of Turkey. Emir. J. Food Agric. typic and genotypic intra-diversity among Anatolian durum wheat 26:149–156. “Kunduru” landraces. Biochem. Syst. Ecol. 65:9–16. doi:10.1016/j. Karagöz, A., and N. Zencirci. 2005. Variation in wheat (Triticum spp.) bse.2016.01.008 landraces from different altitudes of three regions of Turkey. Genet. Baboev, S., A. Morgounov, and H. Muminjanov. 2015. Wheat landra- Resour. Crop Evol. 52:775–785. doi:10.1007/s10722-004-3556-3 ces in farmers’ fields in Uzbekistan: National survey, collection, and Khan, M.K., A. Pandey, S. Choudhary, E.E. Halli, M.S. Akkaya, and conservation, 2010–2015. FAO, Ankara, Turkey. G. Thomas. 2014. From RFLP tp DArT: Molecular tools for wheat Brush, S.B. 2004. Farmers’ bounty: Locating crop diversity in the (Triticum spp.) diversity analysis. Genet. Resour. Crop Evol. 61:1001– contemporary world. Yale Univ. Press, New Haven & London. 1032. doi:10.1007/s10722-014-0114-5 doi:10.12987/yale/9780300100495.001.0001 Koocheki, A., M. Nassiri, S.R. Gliessman, and A. Zarea. 2008. Agro- Brush, S.B., and E. Meng. 1998. Farmers’ valuation and conservation biodiversity of field crops: A case study for Iran. J. Sustain. Agric. of crop genetic resources. Genet. Resour. Crop Evol. 45:139–150. 32:95–122. doi:10.1080/10440040802121445 doi:10.1023/A:1008650819946 Morgounov, A., M. Keser, M. Kan, M. Kucukcongar, S. Dreisigacker, Buerkert, A., M. Oryakhail, A.A. Filatenko, and K. Hammer. 2006. N. Gummadov, H. Muminjanov, F. Ozdemir, D. Sehgal, E. Zuev, Cultivation and taxonomic classification of wheat landraces in the and C. Qualset. 2015. Diversity of wheat landraces currently grown Upper Panjsher Valley of Afghanistan after 23 years of war. Genet. in Turkey: Value for farmers and breeders. Abstracts, 9th Int. Wheat Resour. Crop Evol. 53:91–97. doi:10.1007/s10722-004-0717-3 Conf., Sydney, Australia, p. 37. Carmona, S., J.B. Alvarez, and L. Caballero. 2010. Genetic diversity Negri, V., R. Barocco, L. Pacicco, F. Veronesi, and R. Venanzoni. for morphological traits and seed storage proteins in Spanish rivet 2012. Landraces in Europe: An approach towards identifying land- wheat. Biol. Plant. 54:69–75. doi:10.1007/s10535-010-0010-6 race-rich areas as a priority for protection. In: N. Maxted, M.E. Carvalho, A., J. Lima-Brito, B. Maçãs, and H. Guedes-Pinto. 2009. Dullo, B.V. Ford-Lloyd, L. Frese, J.M. Iriodno, and M.A. Pinheiro Genetic diversity and variation among botanical varieties of old Por- de Carvalho, editors, Agrobiodiversity conservation: Securing the tuguese wheat cultivars revealed by ISSR assays. Biochem. Genet. diversity of crop wild relatives and landraces. CABI, Wallingford, 47:276–294. doi:10.1007/s10528-009-9227-5 UK. p. 118–124. doi:10.1079/9781845938512.0118 Damania, A.B., L. Pecetti, C.O. Qualset, and B.O. Humeid. 1996. Qualset, C.O., A.B. Damania, C.A. Zanatta, and S.B. Brush. 1997. Diversity and geographic distribution of adaptive traits in Triticum Locally based crop plant conservation. In: N. Maxted, B.V. Ford- turgidum L. (durum group) wheat landraces from Turkey. Genet. Lloyd, and J.G. Hawkes, editors, Plant genetic resources conservation: Resour. Crop Evol. 43:409–422. The in-situ approach. Chapman & Hall, London, UK. p. 160–175. Dorofeev, V.F., A.A. Filatenko, and E.F. Migushova. 1979. Wheat. In: Tan, A. 2002. In situ (on farm) conservation of landraces from transitional V.F. Dorofeev and O.N. Korovina, editors, Cultivated flora of the zone of Turkey. Final report for Project No: TOGTAG-2347. Turk- USSR. Kolos, Leningrad. ish Association for Scientific and Technical Research,İ zmir, Turkey. Feldman, M. 2001. Origin of cultivated wheat. In: A. Bonjean and W.J. van Slageren, M.W. 1994. Wild wheats: A monograph of Aegilops L. and Angus, editors, The world wheat book: A history of wheat breeding. Amblyopyrum (Jaub. & Spach) Eig (Poaceae). Wageningen Agricul- Lavoisier Tech & Doc, Paris. p. 3–56. tural University Papers 94-7. Agricultural University, Wageningen, Gökgöl, M. 1935. Turkish wheats, Vol. I. Ministry of Agriculture, The Netherlands. Yesilkoy Seed Breeding Institute Publications No. 7, Devlet Press, Zuev, E., A. Amri, A. Brykova, V. Pyukkenen, and O. Mitrofanova. Istanbul, Turkey (In Turkish). 2013. Atlas of bread wheat (Triticum aestivum L.) genetic diversity Gökgöl, M. 1939. Turkish wheats, Vol. II. Ministry of Agriculture, based on spike and kernel characters. N.I. Vavilov Research Institute Yesilkoy Seed Breeding Institute Publications No. 14, Tan Press, of Plant Industry (VIR). S. Petersburg, Russia. Istanbul, Turkey (In Turkish). Zhukovsky, P.M. 1927. Expedition to West Asia. Proc. Appl. Bot. Giuliani, A., A. Karagöz, and N. Zencirci. 2009. Emmer (Triticum dicoc- Breeding 17:135–158. con) production and market potential in marginal mountainous areas of Turkey. Mt. Res. Dev. 29:220–229. doi:10.1659/mrd.00016

3124 www.crops.org crop science, vol. 56, november–december 2016